DE LA MAREA N D COPPINGER
1068
VOL.
28
Oxidation of N,N-Dialkyl Hydroxylamines with t-Butyl Hydroperoxide. A New Synthesis for Nitrones H. E. DE LA MAREA N D G. M. COPPIKGER Shell Development Company, Emeryuille, California Received September 11, 1961 K,N-Dibenzylhydroxylamine has been oxidized by t-butyl hydroperoxide to the corresponding nitrone in excellent yield. The oxidation of an aliphatic hydroxylamine, N,N-diethylhydroxylamine, has also been carried out. However, in the latter case the nitrone was not isolated as such but trapped in situ to give the isoxazolidine. Chemical evidence and electron magnetic resonance measurements suggest a unique free radical mechanism.
I n the course of mechanistic investigations of aminehydroperoxide interaction, specifically those of dialkylamines, a hydroxylamine intermediate has been considered.’ The fate of such an intermediate has been previously obscured because of the large excess of amine present in such reactions. To settle this question, the oxidations of an N,N-diaralkyl and an N,N-dialkylhydroxylamine have now been examined in a n inert solvent.
demonstrate that a nitrone was formed by treating the crude benzene reaction mixture with 2,4dinitrophenylhydrazine in acid media to confirm that hydrolysis of the reaction product produced acetaldehyde. 0
T
CHa-CH=K-CHz-CH3
I1
Hz0 H
0 II
CHI-C-H
Results
+ CHaCHzNHOH
(2)
N,N-Dibenzylhydroxy1amine.-The reaction of N,NRecent work4 on the addition of nitrones to olefins dibenzylhydroxylamine and t-butyl hydroperoxide was suggested the possibility of trapping the nitrone I1 by carried out in benzene solution (55’). The crystalline reaction with methyl methacrylate to give the isoxaproduct displayed absorption bands in the infrared zolidine, 111. which have been assigned to the nitrone group, 6.29 p , 6.35 p z ; the ultraviolet spectrum also showed a characCH3 0 H teristic band a t 300 mp.a I ‘I I I1
0 -CH=X-
+ CHF=C-C--OCH~
+HaC-C-N-C2H6
The nitrone was hydrolyzed in acid media to yield benzaldehyde which was identified as the 2,4-dinitrophenylhydrazone. 0
t
CaHs-CH=N-CHz-Ce.Hs
I
I
COzCHs
I11
Such a trapping technique permitted the isolation of
Hz0
I11 (mixture of two isomers) in 75% yield. The isolation of 111confirms the structure of the intermediate nitrone as 11.
H
0
II
CaHs-C-H
OH
+ CsHs-CH2-NH I
(1)
Further confirmation of structure I was effected by reduction of the product to dibenzylhydroxylamine with lithium aluminum hydride. Finally the combustion analysis was consistent with the assignment of a nitrone structure (I) to this product. N,N-Diethylhydroxylamine.-N,N-Diethy lhydroxylamine reacted rapidly with t-butyl hydroperoxide in benzene (25’). It was not possible to isolate the nitrone by distillation. However, it was possible to (1) (a) G. M. Coppinger and J. D. Swalen. J . Am. Chem. Soc., 83, 4900 (1961); (b) H.E. De La Mare, J. Ory. Chem., 26, 2114 (1960). (2) L. G. Bellamy, “The Infrared Spectra of Complex hlolecuks,” 2nd ed., John Wiley and Sons, Inc., New York, N. Y..1958, pp. 115,305. (3) M. J. Kamlet and L. A. Kaplan, J. Ore. Chem., 22, 566 (1957). Kamlet and Kaplan report disappearance of the “K band“ (314 t o 352 mp) as a result of a photochemical destruction of the nitrone group. We have observed a similar photochemical destruction of I in this current work. A dilute solution of I in benzene showed t h e following rate of disappearance in ordinary light a t room temperature: Time, hr. 0 24 96
(3)
t
A m mp 0.71 .55 .19
Similar behavior was also noted in dilute isooctane solutions.
Discussion The stoichiometry of the dialkylhydroxylamine oxidation in inert solvent is closely approximated by equation 4, the crude nitrone being isolated in -90% yield in the case of the more easily isolable crystalline nitrone, I. Electron magnetic resonance OH
I
R-CHZ-X-CH~R
+ t-BuOOH + 0
t
R-CH=N-CHlR
+ t-BuOH + HzO
(4)
I. R = CsH5 11. R = CHI
measurements on the reaction indicate that there is an intermediate free radical. The structure of the free radical has been determined to be as follows. (4) (a) R. Grashey, R. Huisnen, and H. Leitermann. Tetrahedron L e f k r s , No. 12, 9 (1960); (h) N. A. Le Bel, J. J. Whang, J. Am. Chem. Soc.. 81, 6334 (1959); (0) C. W. Brown, K. Marsden, M. A. T. Rogers, C. M. B. Taylor, and R. Wright, Proc. Chem. Soc., 254 (1960).
APRIL,1963
OXIDATION OF
X,N-DIALKYL HYDROXYLAMINES
6 --y-5
OH
I
L
OH
+ 1-BuOOH +
]
L
dH
1
r
YH
l+
!R-CHZ-&--CH~K IV6
n -H'
11
HzO
+
&BUTYL HYDROPEROXIDE
1069
amines1 is consistent with the highly nucleophilic character of the former as opposed to the amine.
I
We suggest that the following mechanism is consistent with these observations. It-CHz-N-CH2R
WITH
YT
H R-C=h-CHzR
+ HC
(7)
+H+
v One might expect the oxazirane (V) a s a primary product along with the nitrone, but the best preliminary evidence is that it is not f ~ r m e d . ~ An interesting corollary of our studies of amine-hydroperoxide reactions is the observation that 5,N-dialkyl hydroxylamines react much more rapidly with tbutyl hydroperoxide than the analogous dialkylamines.lb If one assumes that the first step in both reactions is a nucleophilic displacement (equation 5) by amino nitrogen on the 0-0 bond, then this is a clear demonstration of the Pearson-Edwards8 "alpha" effect. At equal molarities (0.6 M ) the reaction of N,X-diethylhydroxylamine and t-butyl hydroperoxide is roughly 35% complete in one hour a t 2.5'. I n contrast, our measurements indicate that the reaction between diethylamine and t-butyl hydroperoxide is exceedingly slow a t 2 5 " ; in fact, no detectable reaction was observed in 120 hours. That hydroxylamines are not important products in oxidations conducted in excess N,N-dialkyl( 5 ) See ref. l a . T h e e.m.r. spectrum consists of fifteen lines, three lines of equal intensity from t h e nitrogen nucleus, each of which is split into B quintet by t h e four a hydrogens. T h e relative intensity a i t h i n each quintet is 1 : 4 : 6 : 4 : 1 . T h e l i n e splittings are AH, 9.1 gauss, .\N, 15.6 gauss. (6) Production of IV could occur by direct electron exchange. Hoaever, the formation of a n intermediate guaternary ammonium salt (,5) b y displacement on the 0-0 bond of t h e hydroperoxide followed b y homolytic cleavage of the N-0 hond seems more consistent with t h e observations of other amine-peroxide systems, C. Walling, "Free Radicals in Solution," John Wiley a n d Sons, Inc., New York, N. Y., 1957,pp. 590-595; C. Walling and N.Indicator, J . A m . Chem. Soc., 80, 5814 (19D8). At any rate it seems certain t h a t the initial step cannot result from thermal deconiposition of t h e hydroperoxide. (7) I n the preparation of I the reaction was run a t room temperature in the dark ([ROzIJl e 0.25 m./l. a n d [RNH*I 0.20 ni./l.) where t h e nitrone hand (6.31 p ) appeared early in t h e reaction and reached a maximum value consistent with t h e disappearance of peroxide. A Rubsequent heating period (50 hr. a t 40-45O) produced only a small increase in the nitrone absorbance. and this was accompanied by a further decrease in the peroxide ab8 irbance. (8) J . 0. Edwards and R. G. Pearson, J. A m . Chem. Soc., 84, 16 (1962).
Experimental Materials. N,N-Dibenzylhydroxy1amine.-Dibenzylhydroxylamine was preparede by heating together 1 mole of benzylchloride, 0.5 mole of hydroxylamine hydrochloride, and 1.5 moles of sodium carbonate in 2 1. of methanol-water (75-25) for a period of 4 hr. The reaction mixture was extracted with ether. The product was recovered from the ether solution and recrystallized from isooctane; yield -50% (m.p. 123-124'). N,N-Diethylhydroxylamine.-Diethylhydroxylamine was purchased from the Aldrich Chemical Co. The crude material was employed in preliminary studies; however, carefully fractionated ~ was employed for materiallo [b.p. 82-83' (108 mm.), n Z 51.41761 the isoxazolidine preparation. &Butyl Hydroperoxide.-&Butyl hydroperoxide of high purit,y was prepared by careful vacuum fractionation of commercial stock (Lucidol). All other chemicals were of reagent grade quality. N,N-Diethylamine.-Diethylamine ( Matheson, Coleman & Bell) was distilled from calcium hydride just prior t o use. Analyses.-Spectral analyses were made on the Perkin-Elmer 21 and the Cary Model 14 spectrophotometers. The Hy-Fi (Wilkens Instrument) flame ionization instrument was used for gas-liquid chromatographic (g.1.c.) analyses. Reaction of N,N-Dibenzylhydroxylamine and &Butyl Hydroperoxide.-N,N-Dibenzylhydroxylamine (7.65 g., 0.037 mole) reacted with t-butyl hydroperoxide (3.
